From newshub.ccs.yorku.ca!ists!torn.onet.on.ca!utgpu!cs.utexas.edu!zaphod.mps.ohio-state.edu!moe.ksu.ksu.edu!ux1.cso.uiuc.edu!uchinews!tira!stephen Tue Jun  9 10:07:08 EDT 1992
Article 6094 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!torn.onet.on.ca!utgpu!cs.utexas.edu!zaphod.mps.ohio-state.edu!moe.ksu.ksu.edu!ux1.cso.uiuc.edu!uchinews!tira!stephen
>From: stephen@estragon.uchicago.edu (Stephen P Spackman)
Subject: Evolution, now [was: Grounding] (no AI here...)
In-Reply-To: christo@psych.toronto.edu's message of Thu, 4 Jun 1992 14:38:25 GMT
Message-ID: <STEPHEN.92Jun5014832@estragon.uchicago.edu>
Lines: 43
Sender: news@uchinews.uchicago.edu (News System)
Organization: University of Chicago CILS
References: <600@trwacs.fp.trw.com> <1992May24.143025.7180@psych.toronto.edu>
	<6924@pkmab.se> <BpBso3.850@psych.toronto.edu>
Date: Fri, 5 Jun 1992 06:48:25 GMT

An important thing to note when contemplating phenotype is that the
long-haul gene-line needs to conserve resources, too. Thus there is
(conventional, Darwinian) pressure (in the very long term - this is,
what, a third-order effect) for genetic encoding schemes that direct
variability into probably-viable distributions. Digital-symbolic
encoding makes this notion meaningful (maybe there IS AI here after
all!), and heterogeneous organism structure (whether intra- or inter-
cellularly) makes its expression as straightforward as anything in
biology ever gets (i.e., not very, but enough that you can get a
handle on the easy cases).

The evolution process can be far more ruthless than at the level of
the individual/species, of course. The resources being conserved are
not things like energy, or even things like herds of animals, but bits
of (internally competing!) encoding technology that make it possible
to have the herds of animals be there. "Be able to make fur" is an
example, and if the way to change the pattern of the fur is to have
all the the instances but a handful die off (and here I'm using
goal-speak, and you should assume the usual caveat that this needs
intepreting in the anthropic sense), that isn't a problem - IF it
works, it hasn't compromised the viability of the strategy.

In this case it's paid for in probability of success though - this is
where the third order "pressure" comes in. A strategy where fur
pattern is easily changed by a dominant gene that can sweep through
the population in a hurry is preferred (is in retrospect more likely
to have survived SOMEHOW, if you prefer) than one where dieback is
part of the long term continuity mechanism.

So another way of looking at "how we got here" is that sometime a LONG
time ago, a metastrategy "neurology flexible in 100,000-year
timeframe" came along and we're one of the applications - the clever
part was getting an encoding for the brain structure that you can dick
around with with mutations, recombinations and sundry exciting
mechanisms and still get an organism that can feed itself.

(Maybe given that substrate, the big brain is a compensation for an
abruptly inadequate sensorium? Pipesmoke.)
----------------------------------------------------------------------
stephen p spackman         Center for Information and Language Studies
stephen@estragon.uchicago.edu                    University of Chicago
----------------------------------------------------------------------
       Believe in Strong AI? I don't even believe in Strong I!


